Google Gemini Explained: Models, Apps, and Features

Google’s flagship suite of generative AI models, apps, and services, known as Gemini, represents the company’s next-generation effort to compete with tools like OpenAI’s ChatGPT and Microsoft’s Copilot [1]. Developed by Google’s AI research labs DeepMind and Google Research, the family includes various models ranging from the large "Ultra" to the offline-capable "Nano," all designed to be natively multimodal [1]. The technology powers both the underlying large language models and the consumer-facing chatbot applications formerly known as Bard [2].

Key takeaways

• Gemini is a family of multimodal AI models developed by DeepMind and Google Research [1].
• The Gemini apps on mobile and web serve as clients for these models, replacing Google Assistant on Android [1][2].
• A paid subscription, Google One AI Premium, unlocks advanced features and integrations with Workspace apps [1].
• Gemini Live allows users to have in-depth voice conversations that can adapt to speech patterns in real-time [1][2].

From Models to Mobile Apps

The Gemini ecosystem is split into the underlying models and the client applications used to access them. The model family includes Gemini Ultra, Pro, Flash, and Nano, with TechCrunch identifying Gemini 2.0 Pro as the current flagship, while TechRadar notes the model is up to version 2.5 Pro [1][2]. These models are trained on text, audio, images, and video, distinguishing them from text-only predecessors like LaMDA [1]. Users interact with these capabilities through the Gemini apps on the web and mobile; on Android, the app replaces Google Assistant and allows users to query content on their screen via an overlay [1]. A feature called Gemini Live enables in-depth voice conversations on mobile devices and Pixel Buds Pro 2, allowing users to interrupt the AI and receive real-time coaching [1][2].

Advanced Features and Workspace Integration

For deeper functionality, Google offers the Google One AI Premium Plan at $20 per month, which unlocks Gemini Advanced [1]. This tier provides access to more sophisticated models within the apps and integrates AI into Workspace tools like Gmail, Docs, and Sheets for tasks such as drafting emails or generating slides [1]. Subscribers also gain access to "Deep Research," which uses advanced reasoning to create multi-step research briefs, and "Gems," which are custom chatbots tailored to specific instructions [1]. Additionally, the paid tier includes features like AI video generation with Veo 2, according to TechRadar [2]. Google also provides a specific experience for teens with additional safety policies and literacy guides [1].

Why it matters

Gemini signifies Google’s push to embed generative AI deeply into its product ecosystem, moving beyond standalone chatbots to become a utility in email, coding, and security [1]. By offering both free and paid tiers with distinct capabilities like reasoning and multimodal interaction, Google aims to maintain its position against rivals like OpenAI and in the rapidly evolving AI landscape [2].